granicus.if.org Git - postgresql/blob - src/backend/parser/keywords.c

   1 /*-------------------------------------------------------------------------
   2  *
   3  * keywords.c
   4  *        lexical token lookup for reserved words in PostgreSQL
   5  *
   6  * Portions Copyright (c) 1996-2002, PostgreSQL Global Development Group
   7  * Portions Copyright (c) 1994, Regents of the University of California
   8  *
   9  *
  10  * IDENTIFICATION
  11  *        $Header: /cvsroot/pgsql/src/backend/parser/keywords.c,v 1.131 2002/11/15 02:50:08 momjian Exp $
  12  *
  13  *-------------------------------------------------------------------------
  14  */
  15 #include "postgres.h"
  16
  17 #include <ctype.h>
  18
  19 #include "nodes/parsenodes.h"
  20 #include "parser/keywords.h"
  21 #include "parser/parse.h"
  22
  23 /* NB: This file is also used by pg_dump. */
  24
  25 /*
  26  * List of (keyword-name, keyword-token-value) pairs.
  27  *
  28  * !!WARNING!!: This list must be sorted, because binary
  29  *               search is used to locate entries.
  30  */
  31 static const ScanKeyword ScanKeywords[] = {
  32         /* name, value */
  33         {"abort", ABORT_TRANS},
  34         {"absolute", ABSOLUTE},
  35         {"access", ACCESS},
  36         {"action", ACTION},
  37         {"add", ADD},
  38         {"after", AFTER},
  39         {"aggregate", AGGREGATE},
  40         {"all", ALL},
  41         {"alter", ALTER},
  42         {"analyse", ANALYSE},           /* British spelling */
  43         {"analyze", ANALYZE},
  44         {"and", AND},
  45         {"any", ANY},
  46         {"as", AS},
  47         {"asc", ASC},
  48         {"assertion", ASSERTION},
  49         {"assignment", ASSIGNMENT},
  50         {"at", AT},
  51         {"authorization", AUTHORIZATION},
  52         {"backward", BACKWARD},
  53         {"before", BEFORE},
  54         {"begin", BEGIN_TRANS},
  55         {"between", BETWEEN},
  56         {"bigint", BIGINT},
  57         {"binary", BINARY},
  58         {"bit", BIT},
  59         {"boolean", BOOLEAN},
  60         {"both", BOTH},
  61         {"by", BY},
  62         {"cache", CACHE},
  63         {"called", CALLED},
  64         {"cascade", CASCADE},
  65         {"case", CASE},
  66         {"cast", CAST},
  67         {"chain", CHAIN},
  68         {"char", CHAR_P},
  69         {"character", CHARACTER},
  70         {"characteristics", CHARACTERISTICS},
  71         {"check", CHECK},
  72         {"checkpoint", CHECKPOINT},
  73         {"class", CLASS},
  74         {"close", CLOSE},
  75         {"cluster", CLUSTER},
  76         {"coalesce", COALESCE},
  77         {"collate", COLLATE},
  78         {"column", COLUMN},
  79         {"comment", COMMENT},
  80         {"commit", COMMIT},
  81         {"committed", COMMITTED},
  82         {"constraint", CONSTRAINT},
  83         {"constraints", CONSTRAINTS},
  84         {"conversion", CONVERSION_P},
  85         {"convert", CONVERT},
  86         {"copy", COPY},
  87         {"create", CREATE},
  88         {"createdb", CREATEDB},
  89         {"createuser", CREATEUSER},
  90         {"cross", CROSS},
  91         {"current_date", CURRENT_DATE},
  92         {"current_time", CURRENT_TIME},
  93         {"current_timestamp", CURRENT_TIMESTAMP},
  94         {"current_user", CURRENT_USER},
  95         {"cursor", CURSOR},
  96         {"cycle", CYCLE},
  97         {"database", DATABASE},
  98         {"day", DAY_P},
  99         {"deallocate", DEALLOCATE},
 100         {"dec", DEC},
 101         {"decimal", DECIMAL},
 102         {"declare", DECLARE},
 103         {"default", DEFAULT},
 104         {"deferrable", DEFERRABLE},
 105         {"deferred", DEFERRED},
 106         {"definer", DEFINER},
 107         {"delete", DELETE_P},
 108         {"delimiter", DELIMITER},
 109         {"delimiters", DELIMITERS},
 110         {"desc", DESC},
 111         {"distinct", DISTINCT},
 112         {"do", DO},
 113         {"domain", DOMAIN_P},
 114         {"double", DOUBLE},
 115         {"drop", DROP},
 116         {"each", EACH},
 117         {"else", ELSE},
 118         {"encoding", ENCODING},
 119         {"encrypted", ENCRYPTED},
 120         {"end", END_TRANS},
 121         {"escape", ESCAPE},
 122         {"except", EXCEPT},
 123         {"exclusive", EXCLUSIVE},
 124         {"execute", EXECUTE},
 125         {"exists", EXISTS},
 126         {"explain", EXPLAIN},
 127         {"external", EXTERNAL},
 128         {"extract", EXTRACT},
 129         {"false", FALSE_P},
 130         {"fetch", FETCH},
 131         {"float", FLOAT_P},
 132         {"for", FOR},
 133         {"force", FORCE},
 134         {"foreign", FOREIGN},
 135         {"forward", FORWARD},
 136         {"freeze", FREEZE},
 137         {"from", FROM},
 138         {"full", FULL},
 139         {"function", FUNCTION},
 140         {"get", GET},
 141         {"global", GLOBAL},
 142         {"grant", GRANT},
 143         {"group", GROUP_P},
 144         {"handler", HANDLER},
 145         {"having", HAVING},
 146         {"hour", HOUR_P},
 147         {"ilike", ILIKE},
 148         {"immediate", IMMEDIATE},
 149         {"immutable", IMMUTABLE},
 150         {"implicit", IMPLICIT_P},
 151         {"in", IN_P},
 152         {"increment", INCREMENT},
 153         {"index", INDEX},
 154         {"inherits", INHERITS},
 155         {"initially", INITIALLY},
 156         {"inner", INNER_P},
 157         {"inout", INOUT},
 158         {"input", INPUT},
 159         {"insensitive", INSENSITIVE},
 160         {"insert", INSERT},
 161         {"instead", INSTEAD},
 162         {"int", INT},
 163         {"integer", INTEGER},
 164         {"intersect", INTERSECT},
 165         {"interval", INTERVAL},
 166         {"into", INTO},
 167         {"invoker", INVOKER},
 168         {"is", IS},
 169         {"isnull", ISNULL},
 170         {"isolation", ISOLATION},
 171         {"join", JOIN},
 172         {"key", KEY},
 173         {"lancompiler", LANCOMPILER},
 174         {"language", LANGUAGE},
 175         {"last", LAST},
 176         {"leading", LEADING},
 177         {"left", LEFT},
 178         {"level", LEVEL},
 179         {"like", LIKE},
 180         {"limit", LIMIT},
 181         {"listen", LISTEN},
 182         {"load", LOAD},
 183         {"local", LOCAL},
 184         {"localtime", LOCALTIME},
 185         {"localtimestamp", LOCALTIMESTAMP},
 186         {"location", LOCATION},
 187         {"lock", LOCK_P},
 188         {"match", MATCH},
 189         {"maxvalue", MAXVALUE},
 190         {"minute", MINUTE_P},
 191         {"minvalue", MINVALUE},
 192         {"mode", MODE},
 193         {"month", MONTH_P},
 194         {"move", MOVE},
 195         {"names", NAMES},
 196         {"national", NATIONAL},
 197         {"natural", NATURAL},
 198         {"nchar", NCHAR},
 199         {"new", NEW},
 200         {"next", NEXT},
 201         {"no", NO},
 202         {"nocreatedb", NOCREATEDB},
 203         {"nocreateuser", NOCREATEUSER},
 204         {"none", NONE},
 205         {"not", NOT},
 206         {"nothing", NOTHING},
 207         {"notify", NOTIFY},
 208         {"notnull", NOTNULL},
 209         {"null", NULL_P},
 210         {"nullif", NULLIF},
 211         {"numeric", NUMERIC},
 212         {"of", OF},
 213         {"off", OFF},
 214         {"offset", OFFSET},
 215         {"oids", OIDS},
 216         {"old", OLD},
 217         {"on", ON},
 218         {"only", ONLY},
 219         {"operator", OPERATOR},
 220         {"option", OPTION},
 221         {"or", OR},
 222         {"order", ORDER},
 223         {"out", OUT_P},
 224         {"outer", OUTER_P},
 225         {"overlaps", OVERLAPS},
 226         {"overlay", OVERLAY},
 227         {"owner", OWNER},
 228         {"partial", PARTIAL},
 229         {"password", PASSWORD},
 230         {"path", PATH_P},
 231         {"pendant", PENDANT},
 232         {"placing", PLACING},
 233         {"position", POSITION},
 234         {"precision", PRECISION},
 235         {"prepare", PREPARE},
 236         {"preserve", PRESERVE},
 237         {"primary", PRIMARY},
 238         {"prior", PRIOR},
 239         {"privileges", PRIVILEGES},
 240         {"procedural", PROCEDURAL},
 241         {"procedure", PROCEDURE},
 242         {"read", READ},
 243         {"real", REAL},
 244         {"recheck", RECHECK},
 245         {"references", REFERENCES},
 246         {"reindex", REINDEX},
 247         {"relative", RELATIVE},
 248         {"rename", RENAME},
 249         {"replace", REPLACE},
 250         {"reset", RESET},
 251         {"restrict", RESTRICT},
 252         {"returns", RETURNS},
 253         {"revoke", REVOKE},
 254         {"right", RIGHT},
 255         {"rollback", ROLLBACK},
 256         {"row", ROW},
 257         {"rows", ROWS},
 258         {"rule", RULE},
 259         {"schema", SCHEMA},
 260         {"scroll", SCROLL},
 261         {"second", SECOND_P},
 262         {"security", SECURITY},
 263         {"select", SELECT},
 264         {"sequence", SEQUENCE},
 265         {"serializable", SERIALIZABLE},
 266         {"session", SESSION},
 267         {"session_user", SESSION_USER},
 268         {"set", SET},
 269         {"setof", SETOF},
 270         {"share", SHARE},
 271         {"show", SHOW},
 272         {"similar", SIMILAR},
 273         {"simple", SIMPLE},
 274         {"smallint", SMALLINT},
 275         {"some", SOME},
 276         {"stable", STABLE},
 277         {"start", START},
 278         {"statement", STATEMENT},
 279         {"statistics", STATISTICS},
 280         {"stdin", STDIN},
 281         {"stdout", STDOUT},
 282         {"storage", STORAGE},
 283         {"strict", STRICT},
 284         {"substring", SUBSTRING},
 285         {"sysid", SYSID},
 286         {"table", TABLE},
 287         {"temp", TEMP},
 288         {"template", TEMPLATE},
 289         {"temporary", TEMPORARY},
 290         {"then", THEN},
 291         {"time", TIME},
 292         {"timestamp", TIMESTAMP},
 293         {"to", TO},
 294         {"toast", TOAST},
 295         {"trailing", TRAILING},
 296         {"transaction", TRANSACTION},
 297         {"treat", TREAT},
 298         {"trigger", TRIGGER},
 299         {"trim", TRIM},
 300         {"true", TRUE_P},
 301         {"truncate", TRUNCATE},
 302         {"trusted", TRUSTED},
 303         {"type", TYPE_P},
 304         {"unencrypted", UNENCRYPTED},
 305         {"union", UNION},
 306         {"unique", UNIQUE},
 307         {"unknown", UNKNOWN},
 308         {"unlisten", UNLISTEN},
 309         {"until", UNTIL},
 310         {"update", UPDATE},
 311         {"usage", USAGE},
 312         {"user", USER},
 313         {"using", USING},
 314         {"vacuum", VACUUM},
 315         {"valid", VALID},
 316         {"validator", VALIDATOR},
 317         {"value", VALUE},
 318         {"values", VALUES},
 319         {"varchar", VARCHAR},
 320         {"varying", VARYING},
 321         {"verbose", VERBOSE},
 322         {"version", VERSION},
 323         {"view", VIEW},
 324         {"volatile", VOLATILE},
 325         {"when", WHEN},
 326         {"where", WHERE},
 327         {"with", WITH},
 328         {"without", WITHOUT},
 329         {"work", WORK},
 330         {"write", WRITE},
 331         {"year", YEAR_P},
 332         {"zone", ZONE},
 333 };
 334
 335 /*
 336  * ScanKeywordLookup - see if a given word is a keyword
 337  *
 338  * Returns a pointer to the ScanKeyword table entry, or NULL if no match.
 339  *
 340  * The match is done case-insensitively.  Note that we deliberately use a
 341  * dumbed-down case conversion that will only translate 'A'-'Z' into 'a'-'z',
 342  * even if we are in a locale where tolower() would produce more or different
 343  * translations.  This is to conform to the SQL99 spec, which says that
 344  * keywords are to be matched in this way even though non-keyword identifiers
 345  * receive a different case-normalization mapping.
 346  */
 347 const ScanKeyword *
 348 ScanKeywordLookup(const char *text)
 349 {
 350         int                     len,
 351                                 i;
 352         char            word[NAMEDATALEN];
 353         const ScanKeyword *low;
 354         const ScanKeyword *high;
 355
 356         len = strlen(text);
 357         /* We assume all keywords are shorter than NAMEDATALEN. */
 358         if (len >= NAMEDATALEN)
 359                 return NULL;
 360
 361         /*
 362          * Apply an ASCII-only downcasing.      We must not use tolower() since it
 363          * may produce the wrong translation in some locales (eg, Turkish),
 364          * and we don't trust isupper() very much either.  In an ASCII-based
 365          * encoding the tests against A and Z are sufficient, but we also
 366          * check isupper() so that we will work correctly under EBCDIC.  The
 367          * actual case conversion step should work for either ASCII or EBCDIC.
 368          */
 369         for (i = 0; i < len; i++)
 370         {
 371                 char            ch = text[i];
 372
 373                 if (ch >= 'A' && ch <= 'Z' && isupper((unsigned char) ch))
 374                         ch += 'a' - 'A';
 375                 word[i] = ch;
 376         }
 377         word[len] = '\0';
 378
 379         /*
 380          * Now do a binary search using plain strcmp() comparison.
 381          */
 382         low = &ScanKeywords[0];
 383         high = endof(ScanKeywords) - 1;
 384         while (low <= high)
 385         {
 386                 const ScanKeyword *middle;
 387                 int                     difference;
 388
 389                 middle = low + (high - low) / 2;
 390                 difference = strcmp(middle->name, word);
 391                 if (difference == 0)
 392                         return middle;
 393                 else if (difference < 0)
 394                         low = middle + 1;
 395                 else
 396                         high = middle - 1;
 397         }
 398
 399         return NULL;
 400 }